Variations in the care of patients with AIDS may contribute to excessive morbidity and mortality and in-hospital costs. Much of the prior research addressing quality of care and resource use in AIDS patients has focused on areas hardest hit by the AIDS epidemic. Using Pneumocystis carinii pneumonia (PCP) as a model for AIDS, we plan to study variations in care of AIDS patients with PCP in a lower prevalence area, Raleigh-Durham, North Carolina, and to compare our findings to those seen in four high AIDS prevalence cities. The principal goals of this project are three-fold: (1) to determine how mortality in AIDS patients who are hospitalized with PCP in the Raleigh-Durham area is affected by: (a) patient characteristics (including age, sex, race, AIDS risk behaviors, health insurance, severity of illness, rural residence, and distance to hospitals), and (b) hospital characteristics (including teaching status, presence of an NIH-sponsored AIDS Clinical Trial Unit, payor mix, and the hospitals' experience with AIDS patients); (2) to compare our findings to those obtained in the ongoing Four-City AIDS Studies which examine variations in care of patients with AIDS-related PCP hospitalized in Los Angeles, Chicago, New York, and Miami; and (3) to evaluate the role of geographic factors such as distance to hospital care or living in a rural area, in predicting access to care, severity of illness, and patient mortality in these North Carolina hospitals as well as in the high prevalence AIDS cities examined in the Four-City AIDS Studies. This study will review the medical records of 300 AIDS patients hospitalized for PCP at any of the 16 non-Veterans Administration (VA) hospitals in the Raleigh-Durham MSA in North Carolina from January 1, 1987 through December 1, 1990. The information will be merged with similar data obtained from the ongoing AHPCR and VA funded 4-City Study of Quality of Care for Patients with HIV-related PCP, creating a Five City AIDS Study database from which analyses may be performed examining both lower and higher AIDS prevalence areas. In addition, this study will use the expertise of researchers within the Rural Health Program at the Cecil G. Sheps Center for Health Services Research to develop and apply methods for evaluating the importance of geographic factors in understanding variations in care for AIDS patients.